Rebalancing costs

Rebalancing sounds free. It is not.

Selling the asset that has run up and buying the one that has lagged means trading. Trading costs money. And selling something that has gone up usually means tax.

So the adviser faces a trade-off, and the workbook states it in one line. It is the cost of rebalancing against the cost of not rebalancing.

The cost of not rebalancing is real too. If the target mix was the right mix, then every drift away from it is unwanted. The portfolio quietly becomes riskier, or duller, than the client agreed to.

The workbook names two kinds of cost. One is transaction cost. The other is tax cost. It then adds a third thing to bear in mind, which is opportunity cost.

How often to rebalance is decided by weighing those against the drift. That decision belongs in the rebalancing policy inside the Investment Policy Statement, not made in a hurry when the drift shows up.

The trade-off (section 15.17.1). Portfolio rebalancing involves a simple trade-off: the cost of rebalancing versus the cost of not rebalancing. While deciding the frequency of rebalancing, that trade-off has to be arrived at. If the investor's target asset allocation is considered optimal, any deviation from it due to price fluctuations is undesirable — which is the cost of doing nothing. But rebalancing costs need to be borne in mind.

The 2 types of cost the workbook names.

1. Transaction cost — the time and money costs of buying and selling securities, which the workbook illustrates with research cost and brokerage. Relatively illiquid investments such as private equity and real estate pose challenges in rebalancing because they carry higher transaction costs. Rebalancing is comparatively easier with more liquid assets such as listed equities or government bonds.
2. Tax cost — while rebalancing, managers sell assets, usually the appreciated asset, and buy usually depreciated assets. Portfolio managers often set price targets for buying and selling. Such actions attract a tax liability, which should also be kept in mind while deciding the frequency of rebalancing and the tolerance for deviations.

And a third item, flagged separately. The workbook adds that there may be opportunity cost as well, with its own illustration: when I move from equity to debt, equity may still do better.

Why the liquidity point matters more than it looks. The two costs do not fall on every portfolio equally. A client whose portfolio is listed equity and G-Secs can rebalance often and cheaply. A client with private equity or direct real estate cannot, so the same 5-percentage-point drift justifies action in one portfolio and not in the other.

The workbook attaches no figure to these costs — no brokerage rate, no tax rate, no tolerance band, no recommended frequency. Section 15.17.1 is qualitative throughout. The percentages in the example below are illustrative and labelled as such.

The workbook gives no cost figures in section 15.17.1; the rates below are assumed so the arithmetic runs.

Mr Bhatt holds Rs 50,00,000 on a 60:40 equity:debt target. After a strong year:

Restoring 60:40 needs equity at Rs 34,80,000, so Rs 3,20,000 of equity must be sold.

The cost of rebalancing. Assume brokerage and other transaction charges of 0.30%, and that Rs 1,20,000 of the Rs 3,20,000 sold is gain, taxed at 12.5%.

The cost of not rebalancing. The portfolio is carrying 5.5 percentage points of unwanted equity — Rs 3,20,000 more equity risk than the client agreed to. If equity falls 25% in a bad year, that excess alone costs Rs 80,000. Against Rs 15,960 of certain cost, the trade is worth making.

Now change only the asset class. Suppose the Rs 3,20,000 of excess sits in a direct real estate holding instead of a listed equity fund. It cannot be trimmed by Rs 3,20,000 at all — property does not divide — and assume transaction costs of 6% on a sale, or Rs 19,200 on that slice even if it could. This is exactly the workbook's point about illiquid assets posing challenges in rebalancing.

And the opportunity cost the workbook names. If Mr Bhatt sells Rs 3,20,000 of equity and equity then rises another 20%, he has given up Rs 64,000 of gain. That is not a mistake. It is the price of holding the risk level the client chose — and the reason the decision belongs in a written policy rather than in a judgement made after the fact.

Chapter 15 (Portfolio Construction Process), section 15.17.1 (Benefits and difficulties of rebalancing), immediately after rebalancing itself at 15.17. Expect a question naming the 2 types of cost — transaction and tax — and one asking which asset classes are harder to rebalance, where the answer is illiquid investments such as private equity and real estate. A subtler question asks what is being traded off, and the answer is the cost of rebalancing against the cost of not rebalancing, not the cost against the expected return.

• Two costs, and the workbook names them: transaction cost and tax cost. Opportunity cost is mentioned separately, as something that may also arise.
• Transaction cost includes time and research, not just brokerage. The workbook lists research cost alongside brokerage.
• The trade-off is against the cost of NOT rebalancing. If the target allocation is optimal, the drift itself is a cost. An option that weighs rebalancing cost against expected return has the wrong comparison.
• Illiquid assets are harder, not cheaper, to rebalance. Private equity and direct real estate carry higher transaction costs; listed equities and government bonds are comparatively easy.
• Rebalancing normally sells the appreciated asset. That is why tax bites — and why a tax-free portfolio can rebalance more often.
• The costs decide the frequency and the tolerance band, not whether to rebalance at all. The workbook puts both questions inside the policy.
• The workbook gives no figures here. Any brokerage rate, tax rate or tolerance percentage in an option is not from this section.